Analysis of Activator-Binding Sites on the APC/C Supports a Cooperative Substrate- Binding Mechanism  Mary E. Matyskiela, David O. Morgan  Molecular Cell  Volume 34, Issue 1, Pages 68-80 (April 2009) DOI: 10.1016/j.molcel.2009.02.027 Copyright © 2009 Elsevier Inc. Terms and Conditions

Figure 1 TPR Groove Mutations in Cdc27 and Cdc23 Cause Cdh1-Binding Defects (A) The TPRs of Cdc27, Cdc23 and Cdc16 are shown in numbered boxes. Residues that we show are important for activator binding are labeled in TPR 5 and TPR6 of Cdc27 and TPR5 of Cdc23. (B) A sequence alignment of TPR 6 from Cdc27 homologs illustrates the TPR consensus residues (green) that mediate the interactions between Helix A and Helix B and are distinct from those predicted to form the surface of the TPR groove (pink). (C) Structure of TPR 3 from the TPR2A domain of Hop (Scheufler et al., 2000), with TPR consensus residues in green and groove-forming residues in pink. This image was generated with MacPyMOL (DeLano, 2007). (D) Cdc27 and Cdc23 mutants have reduced activator binding. APC/C was immunoprecipitated from strains lacking the wild-type subunit and expressing the indicated Cdc27 or Cdc23 mutant subunit. Ten percent of the preparation was used to measure ubiquitination activity with 125I-cyclin B (top panels). A limiting amount of Cdh1 was used to normalize for the amounts of immunopurified APC/C. Ninety percent of the immunopurified APC/C was used for measurement of 35S-Cdh1 binding (middle panels). The gel was stained with Coomassie blue (bottom panels; starred band indicates background IgG from beads) to assess integrity of the APC/C. Three different amounts of wild-type APC/C were tested for comparison (WT lanes). This experiment is representative of three separate experiments, except that the lower amounts of APC/C protein obtained with the Cdc23-K466A and G470A mutants in this experiment were not reproducible. Mutants pursued further are marked with asterisks and labeled. Doc1-4A APC/C was immunopurified using a TAP-tag on Cdc16 rather than Apc1, resulting in the shifted mobility of subunits on the Coomassie-blue-stained gel. Molecular Cell 2009 34, 68-80DOI: (10.1016/j.molcel.2009.02.027) Copyright © 2009 Elsevier Inc. Terms and Conditions

Figure 2 Mutations in Cdc27 and Cdc23 Reduce Cdh1 Response and Activity, but Not Processivity Recombinant 6xHis-Cdh1, purified from baculovirus-infected insect cells, was titrated into ubiquitination reactions containing E1, E2, ATP, ubiquitin, 125I-cyclin B and either wild-type or mutant TAP-purified APC/C. Cdh1 concentrations were 0, 3, 10, 30, 100, 300, 600, and 1000 nM. After 30 min at room temperature, reaction products were analyzed by SDS-PAGE and PhosphorImager (A, C, E, G). These results are representative of three separate experiments. The amounts of mono- (cyclin B-Ub1), di- (cyclin B-Ub2), and tri- (cyclin B-Ub3) ubiquitinated species were quantified and combined to provide the total amount of ubiquitinated cyclin B. The processivity of ubiquitination was determined by calculating the ratio of ubiquitins to cyclin (B, D, F, H). To observe significant activity with the Cdc27-A1A2 mutant, we used 5-fold greater amounts of APC/C, and the quantification reflects the normalized activity per mole of enzyme. Molecular Cell 2009 34, 68-80DOI: (10.1016/j.molcel.2009.02.027) Copyright © 2009 Elsevier Inc. Terms and Conditions

Figure 3 Cdc20-Dependent APC/C Activity Is Defective in Cdc27 and Cdc23 Mutants In Vitro (A) APC/C reactions were performed with either wild-type, Cdc27-A2, Cdc27-A1, or Cdc27Δ APC/C using 35S-securin as substrate and the indicated amount of in vitro translated, ZZ-tagged Cdh1 or Cdc20 purified from rabbit reticulocyte lysates. Control lanes (−) contained mock activator purifications from 5 μl rabbit reticulocyte lysate without activator. Due to the different levels of activity with Cdc20 and Cdh1, it is not clear whether the mutations had a greater effect on Cdc20-dependent activity than Cdh1-dependent activity. (B) APC/C reactions were performed with either wild-type, Cdc23-A or Cdc23Δ APC/C as in (A). Note that in the reaction with wild-type APC/C and 5 μl Cdh1, apparent processivity is increased due to substrate depletion. Molecular Cell 2009 34, 68-80DOI: (10.1016/j.molcel.2009.02.027) Copyright © 2009 Elsevier Inc. Terms and Conditions

Figure 4 Cdc27 and Cdc23 Mutants Have Mild Individual Defects In Vivo and Together Are Synthetically Lethal (A) Log-phase growth rates, at 30°C and 37°C, of strains in which the indicated mutant gene replaced the endogenous gene. Results are representative of three individual experiments. (B) Asynchronous cultures of the indicated cells (AS time point) were treated with α factor (1 μg/ml) for 2.5 hr. α factor was then washed out (zero time point) and cells were harvested at the indicated times. α factor was added back after 90% of the cells were budded. Parallel samples were analyzed directly for budding index (black squares) and fixed, treated with zymolyase, and stained with DAPI to measure chromosome segregation, and with antitubulin antibodies to measure spindle elongation. We counted binucleate cells that had elongated spindles (open circles). Additional samples were prepared for western blotting with anti-Clb2 antibodies and anti-Cdk1 (as a loading control). (C) Asynchronous cultures of the indicated PGAL1-CDC23 strains in galactose-containing media (AS time point) were treated with α factor for 3 hr (α time point), after which dextrose was added for 2 hr to turn off expression of wild-type CDC23. α factor was washed out (zero time point) and cells released into dextrose-containing media. Samples were taken at the indicated times for analysis of budding index, chromosome segregation, spindle formation, and Clb2 levels as in (B). Molecular Cell 2009 34, 68-80DOI: (10.1016/j.molcel.2009.02.027) Copyright © 2009 Elsevier Inc. Terms and Conditions

Figure 5 Analysis of Activator C Box and IR Motif Interactions with Cdc27 and Cdc23 Mutants (A, C, E) Reactions containing 125I-cyclin B and either wild-type or Cdc27-A2 APC/C (A), Cdc27-A1A2 APC/C (C), or Cdc23-A APC/C (E) were performed with in vitro translated wild-type, ΔIR (IR to AA), or ΔC box (IP to AA) Cdh1 in rabbit reticulocyte lysate. Mutant APC/Cs were added at 2- to 3-fold higher amounts than the wild-type. An equivalent amount of lysate without activator was added to control reactions (−). (B, D, F) Quantification of reactions shown in left-hand panels, normalized for background activity (i.e., activity in equivalent (−) lanes was subtracted) and the amount of enzyme added. Molecular Cell 2009 34, 68-80DOI: (10.1016/j.molcel.2009.02.027) Copyright © 2009 Elsevier Inc. Terms and Conditions

Figure 6 Activator Dissociation Rate Is Increased in TPR Mutants and Decreased in the Presence of Substrate (A) Wild-type, Cdc27-A2, or Cdc23-A APC/C was immunopurified and incubated with in vitro translated 35S-Cdh1 for 1 hr at room temperature. After two rapid washes, a 20-fold excess of unlabeled in vitro translated Cdh1 was added as a competitor. Samples were taken at the indicated times to monitor the amount of 35S-Cdh1 that remained bound to the APC/C. When 35S-Cdh1 and unlabeled Cdh1 were added to the APC/C at the same time (premix lane), minimal binding of 35S-Cdh1 was observed. Coomassie staining indicated that the same amount of APC/C was present in each lane. (B) Wild-type APC/C was immunopurified and incubated with in vitro translated Cdh1 for 1 hr at room temperature. Unbound activator was removed and the remaining APC/CCdh1 was incubated with 125I-cyclin B for 1 hr at room temperature. Unbound 125I-cyclin B was removed and an excess of in vitro translated Acm1 (Enquist-Newman et al., 2008) was added as substrate competitor. Samples were taken at the indicated times to monitor the amount of 125I-cyclin B that remained bound to APC/CCdh1. When 125I-cyclin B and Acm1 were added to the APC/C at the same time (premix lane), minimal binding of 125I-cyclin B was observed. Coomassie staining indicated that the same amount of APC/C was present in each lane. A representative experiment is shown on the left, and the graph on the right shows the average of values from three separate experiments (± SEM). (C) Dissociation of Cdh1 from wild-type or Cdc23-A APC/C was measured as in panel (A), except that 50 μM cyclin was included as indicated during Cdh1 binding and dissociation. Coomassie staining indicated that the same amount of APC/C was present in each lane. Molecular Cell 2009 34, 68-80DOI: (10.1016/j.molcel.2009.02.027) Copyright © 2009 Elsevier Inc. Terms and Conditions

Figure 7 Substrate Promotes Activator Binding through a Bivalent Bridging Interaction (A) Cdc23-A APC/C was immunopurified and incubated for 1 hr at room temperature with in vitro translated 35S-Cdh1 and the indicated concentrations of purified cyclin, securin, or Hsl1 fragments. Immunoprecipitates were washed two times for 30 s each before analysis by SDS-PAGE. D box mutant cyclin and D/KEN box mutant Hsl1 (Burton and Solomon, 2001) were also tested. Fold stimulation is the ratio of the amount of 35S-Cdh1 bound in the presence of substrate to the amount bound with no substrate. (B) Dissociation of Cdh1 from wild-type or Doc1-4A APC/C was measured in the absence and presence of 50 μM cyclin as in Figure 6C. Coomassie staining indicated that the same amount of APC/C was present in each lane. Half-times of Cdh1 dissociation were as follows: wild-type APC/C: 20 min; wild-type APC/C plus cyclin: > 200 min; Doc1-4A APC/C: 20 min; Doc1-4A APC/C plus cyclin: 22 min. (C) We propose that activator binds the APC/C at sites on Cdc27 and Cdc23 and shares substrates with a site on the APC/C core during catalysis. Our results suggest that Cdc27 interacts with the IR motif of Cdh1 and Cdc23 interacts with an unknown site on the activator. The bivalent substrate-binding model is explored in detail in Figure S10. Molecular Cell 2009 34, 68-80DOI: (10.1016/j.molcel.2009.02.027) Copyright © 2009 Elsevier Inc. Terms and Conditions